SOLUTIONS
|
|
GritTec Speaker-ID: The mobile client
- the software solution of voice identification
Overview
The software solution of GritTec Speaker-ID: The mobile client (Version 4.10/ Engine 4.00)
is intended for automatic voice identification of a speech signal of unknown voice by comparing with target voice samples of "speaker cards",
existing in the database of system.
This software product supports the high-quality GUI interface and
may be interest for centers of criminalistics, polices, call centers
and the banks which purpose is text independent voice identification
of unknown audio phonogram of telephone negotiations.
Designed algorithm of voice identifications is based on duel comparison
spectra features of unknown voice with the spectra features of
target voice. Spectra features are calculated with provision of
dynamic determinations of channel distortion level and external
hindrances and noises. More
about GritTec's Speaker-ID engine. |
PRICE: 2000$
|
Features
- Operation with low SNR;
- Fast adaptation to changing of channel distortion and external
noises;
- Minimum duration of a speech signal with a voice example used for correct reception of voice parameters for the target speaker - not less 15 seconds;
- Minimum duration of a speech signal with a voice example used for voice identification or voice verification - not less 7 seconds;
- Speaker identification reliability not less than 90% if both of speech signals were recorded in the same channel;
- Speaker identification reliability not less then 85% if both of speech signals were recorded in different channels (cross channels);
- Automatic voice identification or voice verification doesn't require special skills.
Signal requirement
- Signal format: 16-bits linear, Mono channel;
- 8 kHz sampling rate;
- SNR, at least 10 db;
- Frequency range: 300-3400 Hz or better.
Supported signal formats
- Windows PCM, A/mu-Law Wave, Microsoft ADPCM (MS ADPCM), Intel ADPCM (IMA
ADPCM), Microsoft ACM GSM 6.10 (ACM Waveform).
Availability
- MS Windows x86/x64 platforms.
At the moment GritTec offers 7 variants of licensing of this product, each of which can have restrictions on quantity of
speaker cards of target samples:
License model |
Price |
Purchase |
The single permanent license key0, 1, 2, 3 |
License with supporting of 50 speaker cards.
Trial mode is 30 days |
Free |
Purchase Now |
License with supporting of 100 speaker cards.
(and not more 6 voice samples for each speaker card) |
- |
Purchase Now |
License with supporting of unlimited quantity of speaker cards.
(and unlimited voice samples for each speaker card) |
- |
Purchase Now |
The single temporary license key1, 2, 4 with time
limitation - 1(one) year |
License with supporting of unlimited quantity of speaker cards.
(and unlimited voice samples for each speaker card) |
$2000.0 |
|
Where:
1 - We use "Hardware fingerprint" key of client's PC computer for generation of license key, i.e. the
PC computer where client plan to use this program with
current license key. You can found details about "hardware
fingerprint" key in our demo program.
2 - Your purchase entitles for free restoration each your
license key not more than 3 times.
3 - Free upgrades for the entire 1.xx versions and 35% discount for next 2,xx versions.
4 - Free upgrades.
The system supports databases of Paradox format, with a maximum quantity
of target voices in volume no more than 50 000. The end user has
possibility to create and establish new databases with target voices
(see Fig.1).
Fig.1. Preview of speaker card.
Process of input of identified files in the system is organized in such way that allows the end user to get lists of
identified groups, for each of which it's possible to edit the group description. Each group contains the audio files with
unknown voices which in the further will be identified by comparison with database of target samples. The choice of identified
files and their entering in the current identified group may be carried as in a manual mode as in automatic mode, for this
purpose it's necessary to choose an identification source (see Fig.2).
For each the identified file also there is a possibility to enter the description of history of its record, to change
the identification parameters, which is defined by the level of installing the thresholds of probability of errors 1-th
(False Rejection Rate (FRR)) and 2-th (False Acceptance Rate (FAR)) sort (see Fig.2).
Fig.2. Adjustment of identification parameters.
Process of voice identification is executed in an automatic mode, without
post processing of identified files. For this
purpose it is enough to choose the identified group which will be identified, and to start identification process of group
or the current identified file (see Fig.3). After finishing of identification process it's possible to look identification results for
each identified file (see Fig.4).
Fig.3. Process of voice identification.
Fig.4. Start and preview of identification results.
The search form of the target speakers allows to make the automated search of target speakers in the identified files database
(see Fig.5). Search can be made by three requests of the required
speakers, existing in the speaker cards database. The flexible
system of request settings, display forms of search results and
display signal forms of the current identified file and target
sample file allow to make effectively search and studying of the
received request results.
Fig.5. Process of speaker search.
For more information and order of trial key, please download trial demo and contact us via Online Request Form.
|
|
|
|